Globally-Extensive Deep-Time Ocean-Floor Sedimentary Data in Paleocoordinates
Chris Jenkins, INSTAAR, Univ. Colorado Boulder
Abstract
Ocean-floor drillsites, corings, and dredgings represent millions of years of earth and ocean
history distributed along plate tectonic trajectories that may stretch over thousands of
kilometers. To correctly locate the data in their context of climate zones and ocean-basin
structure, they must be transformed to paleocoordinates along the trajectories, back from
the drillsite. This poster describes software to do that transformation on large volumes of
sediment, biological and chemical data, suitable for global-scale statistical data analysis.
Fig. 3. Paleopaths of the
dated deep-time samples,
colored by their present day
(drillsite) plate, all extending
back for 140Ma.
The GPlates software
package (Boyden & others
2011) was used for
reconstructions. A
programmatic (e.g. Python)
interface to Gplates
programmatic rather than a
manual interactive
interfacing is recommended
following this work.
Fig. 1. Illustration of the issue of
paleocoordinates for seafloor
deep-time data.
Data values in the CretaceousRecent data from ODP Hole
1172, East of Tasmania, is
distributed along a trajectory
that spans over 23° of latitude.
Introduction
There is a wealth of data on earth and ocean environments in the age-dated materials of the
ocean drilling programs, and in corings and dredgings from many other expeditions. They are
recorded at their sampling sites of course, but all the deep-time samples will have been
conveyed by plate tectonic motions to that sampling site.
A new combination of software and data allows the such samples to be located to their
paleo-locations – together with their analyzed sedimentary attributes and biologic
components.
Having the data in paleocoordinates in this way – at their original detail, granularity and with
parameter linkages - opens the possibility of large-scale statistical analysis of the sediments
and biota through long periods of earth history. Finer scale phenomena such as ocean
boundary currents and biogeochemical events will be better resolved. Quantitative
assessments of uncertainty will be possible on conjectures about past conditions and events.
And exceptional cases that may indicate something more interesting will stand out.
The marine records from cores, drillings, dredgings are sparse – but constantly being
improved. More can be done to add industry data and word-based data. It will also be
important to add continental platform data to the set. (Unfortunately deformed continental
and oceanic tectonic zones present special problems, yet to be overcome.)
Recovery (m)
Fig. 2. Over 25,000
seafloor cores of
>0.3m length are
described in detail in
dbSEABED, many
with age-dating
information.
Acknowledgements
Thanks to Doug Fils (Ocean Leadership) and Mike Gurnis (Caltech) for discussions on the topic. Particular
thanks to the several institutions which have worked on data accumulation and harmonization: NOAA NGDC
(‘IMLGS’), Columbia University (LDEO ‘IEDA’), JAMSTEC, Consortium for Ocean Leadership, and ODP TAMU
(‘Janus’ database), MARUM (‘PANGAEA’). Their work assisted greatly in the incorporation of huge datasets
into dbSEABED, leading to global integration. Thanks also to the GPLATES project - essential. The Deep Carbon
Observatory (DCO) project of the Carnegie Institution, funded through the Alfred P. Sloan Foundation, kindly
assisted with project funding.
Form of the Input Data
The form of the core sample data is a strong control on the necessary software methods.
• Scientific ocean drilling programs (DSDP, ODP, IODP) have now drilled over 1300 deeppenetration cores into the seabed globally, recovering lithologies of ocean-floor origin
dated back to ~220Ma, Triassic (e.g., von Rad & others 1992). The drillings are somewhat
biased to deep-ocean scenarios. Fortunately, they can be augmented using dated samples
from short cores, dredgings and underwater outcrops (e.g., von Rad & others 1990). The
current total in dbSEABED from all sources is 7,196 dated core, dredge, and drill sites.
• The basic sedimentary-lithologic data accompanying the datings indicate the geologic and
environmental conditions at certain times and locations. This type of data is held in
databases like dbSEABED and GeoMapApp (Goodwilli & others 2010). In the former the
principal source is a ‘top-20’ set of parameters that address sediment textures and
compositions, and also copious data on grain constituents. The total count of dated
samples with geologic attributes is currently 1,033,783. The invaluable initial curation of
these data was done by institutions such as NGDC, TAMU, and JAMSTEC, requiring very
significant time and funding resources.
• It is important to recognize that, in many cases, later phenomena such as consolidation
and diagenesis, tectonics and metamorphism will have been imposed at a subsequent
time and location. Some parameters are strongly affected, especially physical properties.
• Application of age-models to the samples was essential for the ocean drilling datasets, but
was severely limited by two factors. (i) The number of models is very limited. (ii)
Considerable work was required to untangle non-standard sample namings (i.e., leg,
site,hole, mbsf) in the various data sources (databases).
• Most age determinations have significant uncertainty which has several causes: the
availability of age markers (like microfossils), analytical limitations, the breadth of
biostratigraphic units, and crosswalk between biostratigraphy and absolute
chronostratigraphy (Cohen & others 2013). In the current system, based on dbSEABED, an
age is represented as youngest to oldest range with uncertainties quoted above these
central values.
Fig. 4. Distribution of
age-values for
samples contributing
to the datacube,
colored from blue to
red (0 to 100Ma).
The top and lower
panels show presentday and reconstructed
100Ma continental
outlines.
The Processing Pipeline
• Aggregation of the sedimentary data within a unified and globally extensive database
structure. For this project dbSEABED information processing system (Jenkins 1997; Goff &
others 2008) performed the necessary harmonization of the formats, units and
parameters, and integration of the word-based and numerical data types.
• Identification of the present-day plate based on the sampling location. Two mutuallychecking methods were employed, based on the plate Static Geometries of Seton & others
(2012): (i) spherical geometry point-in-polygon; (ii) raster plate depictions. An unresolved
difficulty is the case of sites located in accretionary wedges at consuming plate margins.
• Submitting the collection site information to GPLATES in GPML format and computing the
paleo-coordinate paths at 1my intervals, back to 100Ma. Site identification could not be
carried through the GPLATES process. Positional matching between sampling sites and the
reconstruction seed-points was therefore required.
• Assigning age dates to each sample, preserving their serial order in cores, and associating
uncertainties. This is far from a trivial operation. Matching published age models (Cervato
& others 2005) to data-based site-sample and lab-analysis names is a practical problem.
• Attach the computed paleocoordinates to the samples. Note that the standard
reconstruction path output from GPLATES, in 1Ma steps, is coarse compared to the
chronology of the Quaternary but finer than many age determinations older than
Miocene.
• To visualize the results, paleogeographic basemaps were obtained from the ‘AgeGrids’
collection of Seton & others (2012). For this exploratory study the sediment attribute data
(e.g., carbonate percent) were assigned to gridcells within an XYT ‘datacube’. This is a good
format for data dissemination. In the 2x2dg, 1Ma resolution ‘cube’ 2,694 cells currently
have values.
Fig. 5. Carbonate
values for Ocean
Drilling and other cored
sites, last 100Ma.
The top and lower
panels show presentday and reconstructed
100Ma continental
outlines. The long blue
track (LHS) began in SE
Pacific, is now at the
Peruvian subduction
zone. The long red
track (RHS) started in
the Antarctic
continental massif, is
now at the Lord Howe
Rise.
References
Boyden, J.A. & others 2011. Next-generation plate-tectonic reconstructions using Gplates. In: Geoinformatics:
Cyberinfrastructure for the Solid Earth Sciences, Keller G.R. and Baru, C., eds., Cambridge University Press.
Cervato, C. & others 2005. The CHRONOS System: Geoinformatics for sedimentary geology and paleobiology.
IEEE Data Interop. Proc., 182.
Exon, N.F., Kennett, J.P., and Malone, M.J., 2004. Leg 189 synthesis: Cretaceous–Holocene history of the
Tasmanian Gateway. Proc. ODP, Sci. Results, 189, 1–37.
Goff, J.A. & others, 2008. Seabed mapping and characterization of sediment variability using the usSEABED
data base. Continental Shelf Research, 28(4-5), 614-633.
Gurnis, M., et al., 2012. Plate tectonic reconstructions with continuously closing plates. Computers &
Geosciences, 38, 35-42.
Jenkins, C.J. 1997. Building Offshore Soils Databases. Sea Technology, 38(12), pp. 25-28.
Jenkins, C. & others 2006. Integration of the Stratigraphic Aspects of very large Seafloor Databases using
Information Processing. In: New Techniques in Sediment Core Analysis, Geological Society Special Publns
276, 200pp.
Seton, M., et al., 2012. Global continental and ocean basin reconstructions since 200 Ma. Earth Science
Reviews, 113: 212-270.
Cohen, K.M., Finney, S.C., Gibbard, P.L. & Fan, J.-X. 2013. The ICS International Chronostratigraphic Chart
(updated). Episodes 36: 199-204.
von Rad, U. & others 1990. Mesozoic sedimentary and volcanic rocks dredged from the northern Exmouth
Plateau: petrography and microfacies. BMR Jl Aust. Geol. Geophys. 11:449-476.
von Rad, U. & others 1992. Rift-to-drift history of the wombat plateau, northwest Australia: Triassic to
Tertiary Leg 122 results. Proc. Ocean Drilling Progr., Scientific Results, 122, 765-800.
Goodwillie, A. & others 2010. GeoMapApp: A free, on-line resource for plate margins research and education.
American Geophysical Union, Fall Meeting 2010, abstract ED44A-03.
C. Jenkins, INSTAAR, CU Boulder 19 June 2014